50 BRITISH LEPIDOPTERA. 



with the normal relation to the other appendages and to the meso- and 

 metathorax." In the majority of Psychid female pupae examined, 

 "the lost imaginal parts — wings, legs, antennae, &c. — are evidently 

 represented by confused creases on the corresponding parts of the 

 cuticle." In one unnamed species the wings were distinct " as small 

 pouches on the meso- and metathorax, whilst the legs, antennae, and 

 even the eyes and mouth-parts could be plainly made out, although of 

 a very rudimentary nature." In Notolophus antiqua, the pupal wings, 

 though " very much smaller in the female than in the male, are con- 

 siderably larger than the wings of the female moth." The wings of a 

 female pupa of Hybernla defoliaria are large and well formed and 

 "almost equal to those of the male pupa." The same facts hold for 

 Nyssia zonaria, the " difference between the wings of the pupae of the 

 two sexes being quite inconsiderable." The pupae of the so-called 

 apterous and semiapterous lepidoptera show that the most degenerate 

 species in this direction have been modified from fully- winged forms, 

 and this degeneracy in the imago has been followed by a reduction in 

 the pupal wings, although at a much slower rate. 



The spiracles are not, as a rule, very characteristic or important 

 in the pupa. There are, besides the prothoracic pair (usually buried 

 deeply in the suture between the pro- and mesothorax), eight pairs of 

 abdominal spiracles in the lepidopterous pupa. Of these latter, the 

 first (and often the second) pair is covered by the wings, whilst the 

 last (or eighth) pair is always abortive. 



The cremaster of the lepidopterous pupa is undoubtedly homo- 

 logous with the suranal plate of the larva. Its modifications in 

 various pupae are almost endless, and any definition, unless couched in 

 the most general terms, would fail. It may be a smooth pointed 

 terminal spine, a bunch of prehensile hooks, or a few scattered hooks 

 distributed over the rounded and smooth surface of the anal segment. 

 In some species, " it aids the pupa in working its way out of the earth 

 when the pupa is subterranean ' ' (Packard) , but generally the cremaster 

 is used for the purpose of attachment. In the pupae of cocoon- spinning 

 larvae, belonging to the Incompletae, the armature of curved hooks 

 enables it to retain its hold on the threads of the interior of the cocoon, 

 restraining it "at precisely that degree of emergence from the cocoon 

 that is most desirable " (Chapman), whilst in the suspended or girt 

 pupae of butterflies, the cremaster is the means of attachment to the 

 silken pad to which such pupae are fastened. Packard states that in 

 many of the more generalised moths — Eriocrania, Gracilaria, Prodoxus, 

 Tdntura, Taleporia, Zeuzera, Harrisina, Psychids, Hepialids, &c. — there 

 is no cremaster, although in Tischeria, Taleporia (dorso-anal), and Psy- 

 chids (ventral) two stout terminal spines perform the office of a cremaster. 

 Two similar dorso-anal spines, as well as sundry simple curved setae on 

 the rounded unarmed end of the abdomen, occur in the pupae of the 

 Solenobiids. Packard further says : " In the obtect Lepidoptera, e.g., 

 in the Notodonts, where the cremaster is present, though variable in 

 shape, it may, from disuse, owing to the dense cocoon, be without 

 spines and hooks, e.g., Cerura, or the cremaster may be entirely 

 wanting, e.g., Gluphisia, or only partially developed, e.g., Notodonta. 

 In the butterflies whose pupae are suspended, the cremaster is especially 

 well-developed. Eeference might here be made to the temporary pupal 

 structures in certain generalised moths, which take the place of a 



